Al-Attar H.H. et al, 2013, performed experiments on samples taken from Bu-Hasa Field, which resulted that sulfate concentration has a substantial effect on LSW flooding. From their experiments it was also shown that sulfate beyond an optimum concentration can display a negative impact on the recovery. Based on the experiments performed on a field (H.H. Al-Attar et al, 2013), the optimum sulfate concentration, with a significant oil recovery, is 47ppm. On laboratory scale it can be achieved by diluting SSW by 50 times with DW. Studies of flooding with Low salinity brines with various compositions has been performed in the Laboratory (Hamouda et al, 2014b). But this particular brine composition has been used to assess the field results on laboratory scale.
Diluting the SSW by 50 times with DW was used as secondary injection fluid. Core SG-3 was flooded with SSW as a primary injection fluid and later with LSW as secondary injection fluid. Flooding with LSW 1:50 was also performed to compare the obtained results with the results of LSW 1:10, performed by Hamouda et al, 2014b, in our laboratory. All the results obtained are shown and explained as follows:
From Figure 27 Oil recovery after injection of SSW is 47.4% and increase in recovery from LSW flooding in second phase is 7%, which gives an overall recovery of 54% of OOIP. This Increase in recovery of 7%, after flooding LSW, may be due to a better rock/brine interaction than LSW 1:10, where increase in recovery was~1%. It can also be assessed by looking at the pressure drop measurements.
An initial increase in pressure drop of 1.39 bar after 1.3PV was observed probably due to early water breakthrough. With small fluctuations, dP stabilized at 0.46bar after injection of 2.5PV of SSW at 4PV/day. It is interesting to examine drop in pressure at higher rate of SSW injection.
46 Figure 27: Oil Recovery and dP measured during flooding with SSW and LSW in Core
SG-3.
dP varied from 5.2-4 bar and stabilizes at 4.5bar with a large fluctuation in between. This large fluctuation may be due to large rock-brine interaction and flow restrictions. A drastic increase in recovery of 10% at higher rate gives large indication of rock-brine interaction. As experienced in other experiments pressure drop decreases during the LSW injection. It could be because of low resistance in flow and flow through already swept area. LSW injection at 4PV/day, an initial increase in dP to 0.85bar at 8.5PV was obtained and after 9.3PV dP started stabilizing to 0.83bar with small peaks. But magnitude of fluctuation is much smaller than SSW, because LSW may flow through the channels created during SSW flooding, as in LSW follows the path of least resistant.
While at higher rate dP stabilizes at 3.5bar with large increase in dP of 4.1bar at 13-14PV. Pressure drop values were calculated mathematically also. The calculated pressure drop values for LSW injection are 0.801 bar and 3.43 bar at 4PV/day and 16PV/day, respectively. The difference in pressure drop is approximately of 2%, which is very small. This could be due to error in calibration
0
47 of manometers. It has been stated before LSW injection leads to fines migration. Which leads to increase in pressure drop and diverts the flow of brine to the un-swept area, therefore, decrease the relative permeability for water. Hence, enhancement in the oil recovery.
Figure 28: Oil Recovery, pH of effluents and influents measured during flooding with SSW and LSW in Core SG-3.
pH of effluents was measured for each sample during the process and shown in figure 28.
Lower pH of effluents than the influent was observed in case of SSW injection and a gradual increase to a higher value than influent in case of LSW injection. The maximum pH reached during injection of LSW is recorded at a value of 8.3. These results are similar to what it was observed in the above experiments and observed by other researchers. As for the previous experiments discussed here, the possible explanation for the pH increase is dissolution of calcite minerals. A decrease in pH during injection of LSW at 16PV/day may be due to the equilibrium between rock and brine was reached at this stage of flooding. Hence, there may be less/no dissolution of calcite in the brine.
Oil Recovery (%) pH of Influents pH of Effluents
SSW
48 From the ion concentration results (Figure 29) for the first phase, all the ions have approximately equal concentrations to the SSW. While during LSW 1:50 flooding, all the ion curves except, concentration of HCO3- showed a decrease in slope. The decline rates of the ions for LSW 1:50 of [Na+], [Mg+2], [Ca+2] and [SO42-] are 0.04, 0.0044, 0.00044 and 0.003 mole/L PV. Rate of declination for [SO42-] is 8 times greater than [Ca+2]. This shows lower contribution of Ca2+ in CaSO4 dissolution. Also, it was observed by other researchers (Gomari et al.) that sulfate is consumed in other processes such as adsorption. From Figure 29, [Ca+2] & [SO42-] follows the opposite trend initially (enlarged in circle). This observation was explained by Gomari et al., citing that if adsorption reflects a deficiency of [SO42-], it may mean that chalk surface is continually renewed and dissolution of calcite contributes in stabilizing the calcium ion concentration. Calcite dissolution and alkalinity of the system has also been expressed in form of reactions (10) and (11).
As explained before, pH increases may be due to calcite dissolution, but it does not reflect in the [HCO3-] in Figure 29. The similar results were observed while experimenting with other LSW in the laboratory by Hamouda et al. They also explained that the water exposed to CO2/air to some extent affects the pH and distribution of carbonate species (Hamouda et al., 2014b). This hypothesis is supports the results of [HCO3-] observed in case of LSW (1:50). Deficiency in [Mg+2] could be due to ion exchange between magnesium and calcium which leads to deposition of dolomite/magnesian. Since the maximum pH measured is 8.3, it is very unlikely of brucite {Mg(OH)2} to get deposited because brucite become stable at pH~10.7 (Hamouda et al., 2014b).
Ion exchange between calcium and magnesium expressed in equation (12). Calcium and magnesium ion concentrations reaches its equilibrium at the starting of 11PV. But when LSW injection rate increases, we can see increase in Ca+2 and Mg+2 concentrations (encircled in black circle in figure 29). This may show increase in exchange of Mg+2 and Ca+2 and formation of magnesian.
49
Figure 29: Ion concentrations of effluents relative to SSW taken while flooding with SSW-LSW in core SG-3.
From the experiments performed by Hamouda et al.35, LSW 1:10 was considered as the best case with a recovery of 62%. While after flooding LSW 1:50 total recovery calculated is 54%.
Despite having higher increase in recovery in case of LSW 1:50 (~7%) than LSW 1:10(Hamouda et al, 2014b) (~0.3%), still LSW 1:10 will be considered as the best case.
From figure 30a, we can see that injecting LSW 1:10 at 4PV/day did not show any oil recovery but at 16 PV/day a 0.3% increase in oil recovery was obtained. The similar case was observed during flooding of LSW 1:50, in figure 27. A 7% increase in recovery was observed at
0.0001 0.001 0.01 0.1 1 10 100
0 4 8 12 16
Concentration relative to SSW
Brine PV Injected
Ion Tracking: SG-3 Flooding SSW/LSW (1:50)
Calcium(Ca) Magnesium(Mg) Sodium(Na) Potassium(K) Chloride(Cl) Carbonate(CO3) Sulfate(SO4)
SSW 16PV/D SSW
4PV/D
LSW
4PV/D LSW
16PV/D
50 Figure 30: a) Recovery, pH & b) pressure drop during flooding with SSW/LSW 1:10.
(Hamouda et al., 2014b)
higher rate. This may be explained with the help of pressure drop curves. From figures 27 and 30b, at 4PV/day drop in pressure is almost same ~0.5 bar, which indicates very low resistance in flow.
While at higher rate average dP is 1.5 bar for LSW 1:10 (figure 30b) and for LSW 1:50 average dP is 3.5 bar (figure 27), which is higher than LSW 1:10. This higher dP in case of LSW 1:50 shows higher resistance in flow. Due to this high resistance, oil recovery is higher in case of LSW 1:50.
51 Figure 31: Comparison between experimental (points) and simulated (lines) ion
concentrations (mole/l) for SSW/LSW.
Simulation (using CMG) and experimental ion concentrations are compared in figure 31.
The simulation was also run for the 16PV, though the concentrations reached the equilibrium at 9PV. The decline trend and the concentrations for simulation and experimental data are in good agreement. The difference in starting of decline curves may be due to mixing of brines in experiment.
0.000001 0.00001 0.0001 0.001 0.01 0.1 1
0 4 8 12 16
Concentrations (mole/l)
Brine PV Injected
Simulated Results of flooding SSW/LSW(1:50)
Exp SO4(1:50) Exp Ca(1:50) Exp Mg(1:50) Ca(1:50) SO4(1:50) Mg(1:50)
52